Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[NewtronStar]

Since the 11 of Nov this line has become less stable and can't understand why there is no noise on the telephone side, most of the stats look fine apart from changes in attenuation by 2.0dB after a re sync occurs and they show up as RDI one's.

Did test with a different modem last week Billion 8800NL all that did was cause even more re syncs so went back to ZyXEL 8924 firmware is V18 can't even blame it on XdB as it was steady for 3 months Aug-Oct

May be I'm missing something in the stats that would give me clue as to what it could be.

[Dray]

Are you on MDWS?

[NewtronStar]

Yes have a look under the user of NStar

[Chrysalis]

My gut guess is you getting very shortlived bursts of interference, that are enough to knock out the dsl signal but not enough to register on your data plots.

[Dray]

Looking at the Resyncs drop-down, the resyncs started this new pattern on 9th November

[NewtronStar]

My gut guess is you getting very shortlived bursts of interference, that are enough to knock out the dsl signal but not enough to register on your data plots.

It would help me if I could see these burst on the graph but it's all over and done with in the time it takes for the modem to sync back up, hoping it is not the start of the issue I had last year with Vodafone of course that was put down to being their equipment.

Looking at the Resyncs drop-down, the resyncs started this new pattern on 9th November

That would be the date this has started and nothing here has changed not even a Crimbo tree insight.

[Moderator edited to fix the two quotations.]

[burakkucat]

If it is any consolation, I am also experiencing peculiar "happenings" with my TalkTalk provided service.

The xDSL link will drop out, hence all upper layers are torn down. The device re-trains, the xDSL link comes up and then, within 20 - 30 seconds, drops once again. It was particularly bad yesterday afternoon . . . so much so that I configured the standby VMG1312-B10A and used it in place of the usual VMG1312-B10D. It made absolutely no difference. And that explains why I was so late in attending to the forum, yesterday.

By my testing of the latest release of DSLstats (v6.1) I have seen one oddity -- the US SNRM appears to be at 13 dB, rather than the normal 6 dB.

[NewtronStar]

Consolation is most welcome and never seen or heard of you having any broadband issues on your circuit before in these forums.

My worry with the unexplained resyncs if it goes above 11 in a 24 hour period the DLM will have to do something as long as it's nothing permanent I can deal with it until then.

[NewtronStar]

This retrain is an odd one 20-11-2017 03:39 40000 : 6101 : 5 Teq Dsp
Still unsure why it keeps trying to sync at 40000 kbps.

[DMZ]

By my testing of the latest release of DSLstats (v6.1) I have seen one oddity -- the US SNRM appears to be at 13 dB, rather than the normal 6 dB.

I've noticed this, I'm with TalkTalk too. Not sure what it's indicating but my service seems all ok. (Using the DLink-DSL 3782)

SNR Margin:   12.4dB/6.1dB

[roseway]

I presume this raised upstream SNRM is actually what the modem is reporting, and not a strange bug in DSLstats?

[DMZ]

I presume this raised upstream SNRM is actually what the modem is reporting, and not a strange bug in DSLstats?

Yes for me.
I rebooted my router 2 days ago after applying some settings and prior to that it was 9dB now it's 12dB. It's had no ill affects nor experienced the re-syncs that others reported in this thread. The DS margin remains unaffected. But the raise in US margin indeed I noticed this.

[kitz]

Had alook at your stats and cant see any obvious cause, other than possible SHINE.
You QLN looks at bit noisty at around 3500.  I seem to recall you previously having noise spikes in D1 probably from RFI, but cant recall if you also had any in D3.   
Hlog looks fine, but it is a wee bit more bumpy after ~tone 3500

[NewtronStar]

Thanks for having a look Kitz and was going to predict a resync close to 6pm this evening which happened at 5:41pm there is something out there causing this, also have the PC running DSLstats and this shows up at the resync time but that could be caused by the resync itself 



 

[kitz]

Had another look at todays stats and QLN seems to have calmed down a bit in D3, but Im sure D2 is now a bit noiser than when I looked yesterday.

Your bitswap tends to spike at around the time of the resync.
LETFRS also showing some very minor spiking.
Uncorrected Re-TX also has some spikes.
Both of the above spiked just after midnight this morning as did your bitswap.

I don't know for certain, but for G.INP lines DLM may be looking at LEFTRS.  I did put this question to Openreach but it was kind of glossed over rather than denied. If I was wrong I would have expected a simple no... so just make of that what you wish.

LEFTRS and Uncorrected Re-Tx are kind of an equivalent of CRC/ErrSecs/SES for G.INP lines (but not exactly), so LEFTRS could be something they do monitor, when deciding DLM reversals. 

I can't see any pattern to it, but it does look rather SHINE like.